Reducing antimicrobial use and its impact on the growing threat of antimicrobial resistance will require collaboration of physicians, health care delivery systems, public health officials, and the community. We propose to develop and test physician and patient level interventions in entire communities in a randomized trial to determine if they reduce prescribing and the prevalence of resistance. We will also assess the direct and indirect costs of such a program. The state's two leading managed care organizations, the Massachusetts Department of Public Health (MDPH), and Harvard Medical School will collaborate to achieve the following specific aims: 1. Perform a randomized controlled effectiveness trial in 16 entire communities of an intervention to decrease unnecessary antimicrobial prescribing for children. 2. Measure the effect of the intervention on nasopharyngeal carriage of resistant and susceptible pneumococus and other respiratory pathogens in children. 3. Using targeted active surveillance, measure the intervention's effect on the incidence of both susceptible and resistant serious bacterial infections. 4. Assess the intervention's cost-effectiveness, taking into account the overall costs of medical care, and non-medical care costs for target conditions. 5. Assess physician and patient attitudes about reduced antimicrobial use. Sixteen Massachusetts communities in which most children receive care from a few practices will be randomized to intervention or observation status. We will work through the practices, the MCOs and local institutions to distribute educational materials to all parents in these practices, as well as implement a physician behavior change intervention using credible experts, community opinion leaders, and practice-based peer leaders. We will use MCOs' automated pharmacy and claims data to assess changes in prescribing and utilization of health care, and to provide feedback to prescribers. The prevalence of resistant respiratory flora will be assessed by surveillance cultures from healthy children and the occurrence of resistant bacterial infections will be determined through active surveillance systems in the participating communities. Cost-effectiveness analysis will estimate the financial and other costs per resistant infection avoided.